704 PERIPATUS. 



both in its structure and habits. While abundant moisture is absolutely indispensable 

 to its existence, immersion is quickly fatal ; it is therefore likely to prove of peculiar 

 theoretical value to the zoogeographer since the only conceivable means of transit from 

 one place to another is by land. But the occurrence of distinct genera in Australasia 

 (Australia, Tasmania and New Zealand), in New Britain and in the Malay Peninsula 

 and Sumatra, renders it impossible, in the absence of information concerning the 

 rejsresentation of Peripatus on any of the intervening islands, to construct a scheme of 

 any route of migration in former geological ages, especially as we cannot be certain 

 whether the centre of distribution lay in the Eastern or the Western Hemisphere, 

 although M. Bouvier 1 is of opinion that the American Continent was the original head- 

 quarters of the Onychophoran stock. 



Until quite recently a peculiar correlation has seemed to subsist between three classes 

 of phenomena in their relation to Peripatus, namely: (1) the discontinuity of its areas of 

 distribution ; (2) the generic divergence of the different types ; (3) the method of nutrition 

 of the young in the intra-uterine development. 



It now appears certain that this correlation is neither so simple nor so fixed and 

 definite as had been supposed. If the method of embryonic nutrition, for example, were 

 strictly correlated with the generic divergencies, the selection of any particular method as 

 bearing a more primitive or ancestral stamp than the rest, might well appear arbitrary, 

 since the generic divergence may be taken to depend partly upon the discontinuity of 

 distribution. 



When I first became acquainted with the blastodermic or trophic vesicle of the 

 embryos of the New Britain species {Paraperipatus novae-britanniae) I was completely at 

 a loss to offer any provisional explanation of it based upon morphological principles, but 

 in the course of time I evolved a working hypothesis, an account of which was published". 



Through the kindness of Professor G. B. Howes, F.R.S., I have recently come to the 

 knowledge of an interesting memoir by Dr H. J. Hansen of Copenhagen, on the structure 

 and development of a primitive insect named Hemimerus talpoides, belonging to the order 

 Orthoptera 3 . This insect has been found living in the fur of a West African rat of the 

 genus Cricetomys; it resembles a wingless cockroach in superficial appearance but is more 

 nearly related to the Forficulina than to the Blattina, as is indicated especially by 

 the arrangement of the mouth-parts. The chief peculiarity of Hemimerus lies in its 

 anomalous method of propagation which isolates it from all known insects. It is truly 

 viviparous (not merely ovoviviparous) and in one female Dr Hansen found six embryos 

 at different stages of development, the largest of which, in its convoluted state, measured 

 2"8 mm. in length and the smallest 1"2 mm. 4 



Dr Hansen points out that it is thus evident that Hemimerus gives birth to its 

 young singly and in succession, not simultaneously; moreover the newly born young only 

 differs from the adult in the less number of joints of the antennae and in lacking 

 the sexual modification of the posterior abdominal segments. 



1 Zool. Anz. xxiv. 1901, p. 59. 



- Willey, A., "Trophpblast and Serosa." A contribution to the morphology of the embryonic membranes 

 of insects. Qnart. J. Mia: Sr.., Vol. 41, 1899, pp. 589—609. 



3 Hansen, H. J., "On the structure and habits of Hemimerus talpoides Walk.," Entomol. Tidskr. xv. 

 Heft 1, 1894, pp. 65—94, Pis. 2 and 3. 



4 The length of the parent without the cerci was ll - 5 mm. [Hansen.] 



